Mina Pêcheux


cali rezo's website

PHP, HTML, CSS, JavaScript

One of my many collaborations with the abstract painter Cali Rezo is being her webmaster and maintaining her website. For this, I've created a custom CMS called "Blackboard" in PHP and HTML/CSS/JS to fully control the contents and display of the site.

the caliscope


How would you like to browse paintings in an interactive and intuitive way on your phone? The Caliscope is an online very lightweight web app that lets you swipe through dozens of abstract paintings with only a cube… but thanks to the magic of virtualization, this cube can have more than 6 faces!

collabs with les fées spéciales

HTML, CSS, JavaScript | VueJS | Unity, C#

Throughout the years, I've worked on several projects in collaboration with the French animation studio Les Fées Spéciales:
  • "Shamane" is a software prototype that focuses on data mining for science-based conservation and management of the Przewalski horse. This eco-friendly project is part of the ongoing effort for saving this endangered species!

small tools


Filling your toolbox with various focused standalone programs is a good way to learn more about some programming language or algorithm. So, over the years, I've developed several small tools in Python:
  • Voronoi Image Generator: this mini Python script uses common data viz tools like numpy, scipy and matplotlib to generate images from a randomly generated Voronoi 2D diagram.

scamdoc: ai for scam detection


Each and everyday, we are confronted with new scams and frauds. The web is flooded with more and more strange ads, false job offers or deceitful emails… and what's worse is that authorities are often overwhelmed and unable to react properly. So, it is time we take the matter into our own hands. What if we could automatically predict how dangerous an email or a domain name is? What if you had a simple tool to easily check for scams with a high-rate accuracy?

Anthony Legros and Jean-Baptiste Boisseau are two Frenchmen who have been fighting online scams for years in particular through their participative website www.signal-arnaques.com. They've now added some machine learning to their toolbox to automatically evaluate "digital identities" trust.

During my internship, I was tasked with optimizing two AI models to predict as accurately as possible the chances a given e-mail or domain name had of being a scam. The two models combined are now a webservice called ScamDoc. The proposed tool was to be easy-to-use for users, maintainable for the company and rely on open-source technologies.

The service is absolutely free of charge and can be used as many times as you wish. There is also a pro (non-free) API for companies to access ScamPredictor directly from anywhere.